|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
86.26.61.53
In Reply to: RE: 6C33 OTL posted by Banat on February 18, 2014 at 04:33:52
Dear Banat,
I have the following questions.
1) Since you shared the suggestions of Lew about the circuit posted by Steffen, I do not understand the meaning of connecting R25 and R26 to the other leads of R1 and R2, i.e, at (or practically) at the ground. Any desired negative feedback (NF) should be cancelled. Don't you believe?
2) Do you also share the Lew's suggestion of eliminating completely the NF? In this case, how do you manage with drifts, bandwidth reduction, etc? Could you show measures with/wo. NF?
3) Did you try the minimum acceptable resistance on the output tube cathodes ? Did you measure the difference on the unbalanced idle current during long lasting operation with 0.47 ohm, 1 ohm and 10 ohm?
4) " Single input differential cascoded voltage gain stage ( typical for Atma-Sphere OTL Amp-s) is Much better solution than solution from your presented schematic from many reason ".
You say "...is Much better" in the sense that you listened to a clear upgrading of sound in two circuits with the ONLY difference consisting in the input differential cascoded topology or that in the scheme posted by Steffen?
Many thanks!
Sincerely
Roberto
Follow Ups:
Hi Roberto !1) On Steffen schematic first 6SN7 tube form input differential first gain stage , so connecting two 470K feedback resistor ( R25&R26 )from two balanced output terminals (balanced low impedance output nodes)directly to input tube (high impedance nodes) grids forms balanced global negative voltage feedback loops ,so nothing unusual is there for me .
2)Yes I run my Amps without any applied GNFB , no significant output power stage DC drift ( DC offset) appear , there is no significant bandwidth limit to , except for lowest audio octave , where I made intentionally low frequency roll of from about 20 Hz by adding on each 6c33c cathode 10R W.W.power resistor in parallel with now 2200uF electrolytic condenser .
3)No I did not observed any unbalance at heavy Amp loading with 1 ohm cathodes resistor , but I switch very fast to 10R/1000uF(2200uF) combination with full success , for example my friend Zoki finished two DIY OTL Multiphase Circlotron Amps , each mono block is with 6 x 6c33c-b , as last reliability test his kid was non stop pumping up some very dynamic techno music which includes steady heavy bass lines for around 6 or 8 hours at 2/3 or 3/4 of full output power without of any negative issue ,nominal load(speaker) impedance was 16ohm .
4)Earlier I have try many different input gain stages , some of them was very similar to Steffen input gain stage ( 2x 6sn7 ) but Atma-Sphere differential cascode gain stage was (is) the sonically clear winner .BTW , you should read on the Atma-Sphere site the resource page , there you can find many interesting articles , about global negative feedback influence on Amplifier characteristic ,power paradigm vs voltage paradigm article , and many other articles which altogether form Atma-Sphere OTL amplifier design philosophy .
Best Regards !
__
Enlightened Evolution-Astral Projection
Edits: 09/08/14 09/08/14 09/08/14
Dear Banat,
many thanks for you kind support: I go to read the pages you have indicated, and I appreciate your information about audio tests of the cascode versus standard differential input stages.
Going back to the example of the design of Steffen, are you thus saying that no drift and any bandwidth reduction would occur if I simply eliminated R25 and R26?
Sincerely
Roberto
Roberto- Definitely No DC output drift will occur if you disconnect that two GNFB resistor , since that OTL amp is Not internally DC coupled , but is AC coupled , first time in between first and second voltage gain stage (C1/C2 ), and second time in between second voltage gain stage and CF (cathode follower) driver stage (C3/C4).
- but some effects you can expect if you totally eliminate GNFB , Amp overall voltage gain will rise up , max. THD % to , Amp internal impedance will rise up to , maybe some hum or noise will appear to ,
Actually I think that you need first to build up that Amp and than to make some listening test ,first with GNFB applied and than without GNFB ,later you can chose variant which you personally prefer more , but for sure that some interesting sonic surprises is expecting you in both cases :)BTW , there is eternal clash between various opinions in audio world about GNFB , some designers insist on no GNFB amplification , others are in middle position saying that moderate applied GNFB is just the right approach , and some of them insist on maximum as possible inserted GNFB , for example well known designer Bruno Putzeys have presented his view in the article named `The F-word `.
__
Enlightened Evolution-Astral Projection
Edits: 09/08/14
Dear Banat,
you are so kind giving me all these suggestions.
Actually, I did not pay attention on these DC blocks, as I had in mind the similar G20 GRAAF schematics (which may be was circulating in this website) and I assumed the connections between stages were in DC as in the GRAAF.
Now I have the point. Many thanks again!
Sincerely
Roberto
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: